OBJECTIVES:

• Hands - on Director of Research & Development in large firms
• Technical partner in small start-ups.

SUMMARY:

• Hands-on skills in computer science, mathematics and quant finance.
• 12+ years of development & leadership with track record of successfully developing quantitative and technical systems.

SKILLS:

Computer: Python (12 years), C++/STL/Boost/TR1 (13 years), Unix/Linux (14 years), Oracle/MySQL/Sybase (13 years), VBA/Excel (12 years), Core Java (12 years), Scala, Matlab, R, Mathematica and experience in many other technologies.

Math: Numerical recipes, PDE, Monte Carlo simulation, statistics, probability, finite difference method, stochastic calculus, binomial tree, linear algebra, regression, optimization, Markov model, neural network, time series, Kalman filter, genetic algorithm and programming, calibration techniques.

Finance: Equities, Derivatives, Black-Scholes model, volatility surface, Greeks, exotics, FX options, fixed income/interest rate derivatives such as bond, swaps, swaptions, cap/floor, Bloomberg, Reuters MarketQA and FASTICK, etc.

EXPERIENCES:

Confidential

Lead Developer & Architect

Responsibilities:

• Developed the cross-asset derivatives market/credit risk management system in Python/C++/Java. Built a global dev/qa/support team and coordinated with several departments on the “Quartz” project, BOA’s risk management infrastructure.
• Built DataSynapse grid computing libraries in Python, implemented concurrent instrument loaders and integrated pricing models for equities, equity derivatives, equity swaps, FX options/futures and fixed income such as treasury bonds and corporate bonds and integrated the derivatives C++ pricing models into the python environment. Implemented the job scheduler in python.
• Built a bitemporal reference data platform (the golden source) for fixed income, bond ratings, interest rate derivatives, equities, equity derivatives, and FX options.
• Started from a party/account reference data system on MongoDB, later migrated to an in-house object-oriented distributed NoSQL database and planed to cover all instruments.
• Built adaptors, C++/Java wrappers, XML data feeders/parsers, XSLT transformers, SOAP clients, TIBCO RV messaging queues to integrate legacy existing systems and consolidated data into one platform.
• Built various APIs and implemented flask RESTful web services and Excel plug-ins. Index search, parallel brute-force search and cross-reference mapping and support JSON, XML and table formats.
• Built Django/jQuery based object browser and dashboard tools for all reference data python objects, XML request/response, instrument query by IDs, data quality analysis, daily unit tests report, RDF reconciliation report, statistic data charting and real-time job status monitors.
• Provided training sessions on Python and system architecture.

Confidential

Team Lead

Responsibilities:

• Recruited and led a team to build a research platform for equity and option strategies on trading floor and report to fund manager. This platform consists of a calculation library, equity derivatives pricing models, linear/nonlinear optimizer, portfolio construction, CAPM, Fama-French 3 factor model, tracking error, Barra risk model, T cost model, simulator and time series plot utilities. Integrate with Market QA, Snopt, R, Matlab and database. Implement from scratch with Python (numpy/scipy/PyTables/pandas), mongoDB, C++, Java, and Fortran
• Machine learning, data modeling, long/short strategy research and model implementation. Research optimization algorithm and optimal execution, high frequency tick data, fastick API. Implemented filters to remove outliers and noise. Large intraday/tick by tick data sets parallel handling in HDF5, Python scikit-learn, C++.
• Solve all kinds of issues for traders by providing quick and immediate answers as well as long term solutions.

Confidential

Senior Quant/Calculation Engineer

Responsibilities:

• Coordinated with traders, IT, quants and reported to CTO. Developed a calculation library for the emerging market interest rate derivatives pricing system.
• Prototyped in Python, VBA/Excel rapid application, Matlab, Core Java, C++ pricing server, QuantLib Java wrapper, Boost multithreading, JNI/RMI middle layer, Sybase database, linear optimization, numerical methods.
• Derived and implemented a creative approach to fast calculate currency swap forward rate and convexity adjustment matrix and time series within any time span. Validate the swap rate with the forward rate from swaption models. Calibrate yield curve for emerging market currencies.
• Prototyped the pricing models for swaps, swaptions, caps and floors. Prototyped Monte-Carlo implementation of two-factor Hull-White model for path-dependent interest rate derivatives on GPU PyCUDA.
• Converted implied volatility between BPvol and yield vol by using lognormal and normal model. Cap/floor volatility conversion. Swaption volatility conversion. Modified Java Apache’s Brent method to solve extreme scenarios.
• Calibrated smooth interest rate curve from Brazilian interest rate future options. Calculated BRL and MXN forward rate matrix.
• Created flat yield bond model to calculate convexity adjustment time series. It is significantly faster than the old method of shifting the IR curve up and down to calculate the real convexity.
• Researched and developed the theoretical butterfly method to construct volatility surface. Improved the existing Risk Reversal pricing and volatility surface construction by introducing a new concept “theoretical butterfly”.
• Researched and developed the average strikes method to construct volatility surface. Our old models (based on vanna-volga method) have significant arbitrage violation and experienced traders have to adjust the strategies’ price by guessing. A new method was constructed to rebuild consistent volatility surface for all market scenarios, especially skewed market which would be better than vanna-volga method and other calibration schemes.
• Developed the binomial options pricing model to support arbitrary payoffs. Implemented both CRR method and the equal probabilities method. Researched the convergence pattern and made the results more accurate for any specific number of time-steps.
• Developed the trading system and pricing server. Added new features, integrate new models, Quick FIX, VeriFIX, and rewrote legacy modules, TCP/IP, socket, multithreading.

Confidential

Lead Developer

Responsibilities:

• Coordinated with several departments to implement critical derivatives pricing solutions for the trading business. (C++, R, Python, VBA/Excel, multithreading, Matlab, Mathematica, Django)
• Developed trading system for equity swaps, equity and index FLEX options.
• Setup a Linux cluster and built a python job scheduler and distribution systems for mark-to-market value calculation and trades processing. Built Django web portal for job viewer and management.
• Designed and implemented grid computing scheme for Monte Carlo simulation on Linux cluster.
• Replaced Monte Carlo with Crank-Nicolson method for double barrier option pricing. I also improved the interpolation method to avoid negative volatility in extreme case, which always fails the Crank-Nicolson method. Integrated into pricing server.
• Data Analytics and Visualization. Designed and implemented an interactive visualization tool for the internal data format of Spotfire financial analytics toolkit.